The impact of different response alternatives on responders' reporting of health-related behaviour in a postal survey

Previous experimental research in other topic areas has shown that the choi ce of response alternatives can influence respondents' reporting of the fre quency of vaguely defined events and that the set of response alternatives is treated as information in the interpretation of the question. The aim of this study was to examine whether such affects would occur in the context of respondents reporting of health-related events using high and medium fre quency closed format response categories, which might be used interchangeab ly by researchers. The study consisted of a postal survey of n = 518 patien ts aged greater than or equal to 18 years randomly selected from the patien t list of a diabetes centre and who were equally and randomly allocated to one of three conditions (Condition A: high frequency response alternatives/ horizontal orientation; condition B: medium frequency response alternatives /horizontal orientations; condition C: high frequency response alternatives /vertical orientation). Testing for the effect of response alternatives for the combined responses of five vaguely defined questions between condition s A and B was chi (2) = 5.5, p = 0.019, for the difference in proportions, indicating that overall, those respondents presented with response alternat ives discriminating at medium frequency, reported significantly fewer targe t events than those presented with high frequency response alternatives. Te sting for the effect of orientation of the combined question responses betw een conditions A and C, differences in proportions between conditions, did not reach statistical significance (p > 0.05). Findings from this and previ ous studies indicate that response alternatives provide information on the interpretation of vaguely defined questionnaire items and that their choice should not be left to intuition alone when designing questionnaire items.